But how does it work? How does it determine which model to use? Is it highly effective at making that determination, or is it a feature that improves over time?
I’d prefer to wait.
They’ve always prioritized using more cost-effective models in their UI to some extent.
The AUTO feature will likely consider OpenAI’s interests as a stakeholder, rather than focusing solely on the users’ best interests.
This is frustrating. It deleted my chat history for that conversation, and I had important information in it. Avoid using it if you want to retain the full chat history.
Yeah, I wouldn’t bother with the lower models. Plan your questions and wait for your turn.
Each release brings a significant boost in capability, so I’d prefer not to have my request handled by a lesser model.
true that’s a good point! makes sense to me that they definitely want to keep the convo as cheap as possible with that, I wonder if it would ever consider using the o1-preview model (which would be the most expensive cost, right?)